A study was conducted to examine sources of variation introduced into a phagocytosis assay as a result of the isolation of neutrophils from bovine blood, including variation attributable to isolation of neutrophils from blood, variation between duplicate determinations of percentage phagocytosis, and the variation in the ability of neutrophils isolated from blood (over repeated collections from the jugular vein) to phagocytose. For the phagocytosis assay, jugular venous blood from each of 4 cows was divided into 2 equal portions. The neutrophils were isolated by lysis of red blood cells with 0.2% sodium chloride. The neutrophils (2 X 10(7)) were incubated in duplicate with 32P-labeled Staphylococcus aureus ([32P]SA; 2 X 10(8)) inskimmed milk samples (2.5% final concentration) prepared from 4 cows. This process was repeated thrice on neutrophils isolated from 4 cows at 2-week intervals. The proportions of variation in percentage of 32P-labeled S aureus phagocytosed between duplicate neutrophil isolations and between duplicate assay determinations were 0 and 1%. Differences among skimmed milk sources and among runs, using blood neutrophils taken at different times from the same donor cow, accounted for 62 and 36% of the total variation. The results indicated that variation arising from blood neutrophil isolation introduced into a phagocytosis assay within a single-day trial is of no concern. The large variation among skimmed milk sample sources indicated differences among cows in the ability of their milk to support phagocytosis. The variation in neutrophil isolations over time for any cow was considered too large to allow for evaluation of physiologic and environmental effects on phagocytosis of neutrophils isolated from blood.